Re: Definition of belief again (was Re: virus:Other Reality)

Reed Konsler (
Wed, 15 May 1996 00:50:05 -0400

Maybe we've lost sight of the original goal of this thread which is to figure
out what is meant by the term "belief". I don't think existence or observation
should be part of the definition any more than for any other noun in the

That's a good point. But it immediately raises the issue we have been
grappling with: ontology.

Logical behaviorism faces similar difficulties in translating
sentences about (what Russel called) propositional attitudes (i.e.,
beliefs that p, desires that p, hopes that p, intentions that p,
and the like). Consider the following sample proposal: one believes
that the cat is on the mat if and only if one is disposed to assent
to 'The cat is on the mat'. First the proposed translation meets the
condition of being purely behavioral only if assenting is understanding
purely in behavioral terms. That is doubtful. The proposal also fails
to provide either a sufficient of necessary condition for believing
the cat is on the mat. Someone may assent to 'The cat is on the mat'
and yet not believe the cat is on the mat; for the person may be
trying to deceive or may have misunderstood the sentence. A belief
that the cat is on the mat will dispose one to assent to 'The cat is
on the mat' only if one understands what is being asked, wants to
indicate that one believes the cat is on the mat, and so on. But none
of these conditions is required for believing that the cat is on the

The point, as I see it, is again in trying to argue for absolutes. Everything
one knows is infered (including the base reality we all seem to share) thus one
can only categorize or define things provisionally.

This gets back to the thought experiment of the actors play-robbing the bank
(was that from Stephen? , I've lost the clip). That scenerio had several
different points-of-view.

An individual is behaving rationally if they act rationally based upon their
point-of-view. They might be dead wrong. Reason does not specify truth in an
instantaneous sense (or else we wouldn't have these long winded discussions);
it allows truth to be approximated by an iterative process.

I don't know how do resolve this one. We could define things based upon an
idealized perspective. My gut tells me that an omnicient perspective is
inherently paradoxical, has anyone read anthing good on this topic? It's a
compromise I'd rather avoid, in any case.

Thanks for keeping us on track, David. It's nice of you to bring this up again
so that I can conceed that your definition IS (as you pointed out earlier) more

"To believe X is to act as if X is true"

I'm still troubled by the implied perspective of this statement, but I'll
accept it provisionally.

William Poudstone, in Labyriths of Reason, brings up an interesting concept in
his description of the Raven dilemma. Maybe we can apply it here:

Assume you had a Robot which you were going to use as a sort of belief
census-taker. What sorts of instructions would you give the Robot so that it
could accurately assess the beliefs of someone else, assuming it's powers of
perception are near-human?

Speaking of Robots: does a computer "believe" it's program? A computer acts
as if the program's instructions were true, or else it couldn't add 2 and 2 to
get 4. Is there a component to human belief that it contains a bit of doubt,
by definition...or at least a little flexibility?